Just when you think you've seen it all, Yamaha come straight at you out of left field with an instrument unlike any other. But is the Tenori-On a musical revolution in the making or a white (flashing) elephant?
Designed by Japanese media artist Toshio Iwai, the Tenori-On is conceived as a new musical instrument for the digital age. Continuing from where his Electroplankton software (for the Nintendo DS) left off and adding a generous dollop of step-sequencer immediacy, Iwai marries visual appeal with an integral synthesizer — and alarm clock! The much-heralded pilot release has been UK-only, suggesting that this is uncharted territory for all concerned.
Resembling an object from the Starship Enterprise crèche, this is a quite unprecedented hi-tech cyber-toy. Yes, there are parallels with existing devices such as the Monome or even the Genoqs Octopus, but unlike the former, the Tenori-On requires no computer to perform its tricks, while comparisons with the latter are primarily due to its LED grid method of note entry.
The Tenori-On is small, approximately eight inches square, and constructed of rounded, polished magnesium. It rattles slightly when handled, courtesy of the myriad white buttons forming the user interface and animated display. There are 256 of these buttons, with an equal number on the rear, although those on the rear are display-only. From the outset, it's clear that the Tenori-On is as much about visuals as it is about sound; its graphical shenanigans are presented to both performer and audience, assuming it is held vertically (which isn't actually the most comfortable playing angle) rather than flat on your lap.
Power is supplied by an external adaptor or by six AA batteries. Battery power offers the tempting prospect of performance in any environment, and its tiny 1W stereo speakers proved adequate for close-in listening at least. For more intimate music-making, a mini stereo headphone socket is provided; this serves as a line out should you wish to plumb into a more substantial sound system. The remaining port to the outside world is a Mini DIN socket, providing MIDI In and Out via a breakout cable, so you can synchronise playback with other Tenori-Ons or talk to external MIDI gear.
With all connections made along the base of the instrument, the wires do tend to get in the way. It's hard to imagine where else they could have gone without a total redesign, but sometimes handling felt a little awkward. Top of my wants list would be a stand or cradle to safeguard the connections during play.
By far the most eye-catching aspect of the Tenori-On is its 16x16 matrix of opaque plastic buttons. Serving as both note and data entry points, they are backlit in three discrete intensities, resulting in a striking animated display. It's worth checking out some of the on-line video demos to appreciate the impact of this, but even so, the reality is better. In addition, the user interface offers five buttons to be operated with the left hand (L1 to L5) and five more for the right (R1 to R5). Add a small backlit display, a jog wheel and two buttons marked OK and Cancel and that's almost it. There's one small button you might miss, right at the top: Clear. This is your means of instantly wiping notes you have entered on the panel.
My final observation before I commence making noises, is that, although the buttons seem robust enough to withstand careful finger prodding, I treated them with respect throughout, as I doubt they would prove inexpensive to replace. Were I to own one of these, I'd definitely invest in some sort of case or bag to protect it.
The Tenori-On generates notes using six different methods, known as Modes, distributed unequally across 16 Layers. Some of the modes offer unique ways to create patterns of notes, so we'll look at each one in turn. For now, suffice it to say that the first seven layers generate notes using Score Mode whilst layers 8 to 11 make use of Random Mode. Layers 12 and 13 are assigned to Draw Mode. Layers 14 and 15 are Bounce and Push Modes respectively. This leaves the final layer assigned to Solo Mode. The mode-to-layer assignments are fixed, and initially there's a lot to take in — but you can explore the capabilities of each in turn, and after a short time, I guarantee you will be leaping from layer to layer without even thinking.
Each layer is assigned to play any of the 256 onboard patches. Thus layer 1 might play strings, layer 2 could be bass, layer 3 drums and so on. As well as patch selection, each layer has its own level and panning, along with speed, transposition and note length.
Rather than spend any longer setting the scene, let's see how this all works for real. The Tenori-On cries out to be played and it makes far more sense to approach it this way. Having powered up, we select a layer to work on courtesy of the R1 button. Hold this and the current layer is represented as a horizontal row of lit LEDs; select a new one by hitting a button on another row. We'll stick with layer 1 for now. Hitting OK starts playback, and four white LEDs march across the display to indicate which step is the current one, as on a typical analogue-style sequencer.
Layer 1 operates in Score Mode, which operates in a similar fashion to high-end hardware sequencers such as the Genoqs Octopus, allowing speedy creation of looping patterns. Each button on the horizontal axis represents a note, with chords achieved by activating notes in a vertical column. Briefly pushing any button triggers its note and an animated pattern of light splashes out briefly from it. Push the button for slightly longer and it remains lit — then, whenever the chase LEDs pass that step, the note and animation are triggered.
Now is a good time to be fully introduced to those vital 'L' and 'R' buttons. We've seen that R1 is used to select the current layer, so we'll quickly run through the others. R2 is used to set the tempo — you can either hold down R2 and touch any of the matrix buttons, giving a range from 50-200 bpm in 10 bpm increments, or you can flick the jog wheel with your thumb, for precise tempi. The jog wheel is used in this way consistently, which is nice. R3 is the master transpose button, its values set in semitones. R4 sets the volume of each layer and R5 is used to select a new block (more about blocks later).
The left-hand buttons are the means of adjusting aspects of the current layer. Thus, L1 selects a new sound, referred to as a Voice. L2 alters voice length from 10ms up to 9990ms, whilst L3 shifts the octave. Five octaves of transposition are available, up or down, although low notes aren't always easy to hear through the built-in speakers. The function of L4 differs according to the mode; as we are currently exploring Score Mode, it is used to set the loop start and end points. A Score Mode loop may be from one to 16 steps — so if you create patterns on several layers at once, each with a different loop length, you are effortlessly transported into a world of complex polyrhythms. Finally, L5 changes layer speed — so a layer can play twice as fast or twice as slow and so on.
Within no time, I was tapping notes into the button matrix and making speedy tweaks using the L and R buttons. The Tenori-On is a delightfully inclusive instrument and I'm sure anyone could get into it, with or without musical knowledge.
The jog wheel and OK/Cancel buttons are used to navigate the menu system, which offers a number of functions for which there is no direct access. There are menus for file operations — for saving and loading songs, blocks or individual layers — and a menu option for setting defaults, such as the preferred sounds or animations for each layer. Menus allow the wiping of a block or layer and, unlike using the Clear button for the same purpose, offer a welcome Undo.
Sweeping global changes can be made, such as setting a master loop length and timing — this overrides those parameters set within individual layers. You can mute the overall output (handy as there is no hardware volume control), reset the loop playback and even prime the integral alarm clock, ready for breakfast.
Rather than list every option, I'll quickly mention the two I found of most value. The first of these I've referred to elsewhere and is found within the Play menu — the Scale selection. The scale affects the entire instrument, except any layers using drum kits or user voices, and there are nine scales on offer. As the Tenori-On neatly avoids details of note names or root, its scales are actually modes (if I remember my boyhood piano lessons correctly). Anyway, the usual modes are present, including ionian (major), chromatic (all notes), aelioan, mixolydian and so forth. It's a pity you can't select a different scale per block or define your own scales (as on, for example, the Sequentix P3) but this feature does simplify music-making nicely.
The final noteworthy menu-based feature is 'Record'. This is a real-time means of capturing a complete song performance, direct to an SD card. As far as I can tell, it includes every button press and parameter tweak. When you're done, you can pass your songs around your circle of Tenori-On-owning friends.
Although you can achieve much with multiple layers of Score Mode, let's add a fresh layer to the mix. We'll choose layer 8 because it's the first one that uses Random Mode. Here you can achieve something other than robotic repetition. Push a button and a single note is repeated at the current layer speed. Select a second one and the two notes alternate, the time between repetitions proportional to the distance between them. Successive notes are played back in the order entered, and play direction is shown as an LED in motion. The results obtained are amiably unpredictable and made even more so when the L4 key is brought into play. In Random Mode, L4 becomes 'rotate': hold it down and draw a direction of rotation on the button matrix with your finger. Then on the start of each bar, your note pattern shifts around, resulting in different notes being played for each passs. And if some of those notes are close to the edge of the grid, rotation may cause them to disappear temporarily, adding to the fun.
Each of the other modes offers its own idiosyncracies. Draw Mode is rather tricky to control at first and I found myself frequently needing the Clear button. In this mode you play the buttons as if on a keyboard, layering and overdubbing as you go. As you can't always predict which button corresponds to the note you want, I tended to slow down playback to maybe half or a quarter of normal speed to increase my guess time. Draw Mode is ideal for tapping in percussive loops or generating crescendos, in sparkling whirls of sound and light.
The next mode, Bounce Mode, is highly entertaining. Touch a matrix button and it's as if you release a bouncing ball which, when it reaches the bottom of the display, triggers the relevant note or percussion voice. In this mode you have to imagine the keyboard laid out horizontally; the closer to the bottom you release the note, the faster it repeats. Whilst compelling to watch, I can understand why there's only one layer dedicated to this mode — it isn't especially controllable. To stop each note bouncing, hit the lowest key in that column.
Push Mode is intended for sustained, pad-type patches. Here notes are set by pushing and holding buttons. The manual recommends the patches in the second row of 16 as being ideal for this purpose because they are programmed to evolve. I'd have loved several Push Mode layers because creating ambient, shifting soundscapes is something the Tenori-On does rather well.
Last up is Layer 16's Solo Mode. This layer doesn't store note values, it is reserved for manual performance. Hold and sustain a single note and it repeats; you change the repeat interval by sliding your finger down to a lower note.
All the various modes generate animations in conjunction with notes — squares, circles, crosses and so on, pulsing, throbbing and winking merrily. These definitely enhance the Tenori-On experience and you can switch them during playback, change their size or determine whether they grow or shrink when activated. It means that any layer with special musical significance can be emphasised visually — although it does get chaotic when clusters of notes are triggered simultaneously on multiple layers. But where pretty flashing lights are concerned, you're never going to hear any serious criticisms from me!
Having built a hypnotic pattern with as many active layers as necessary, you will eventually want to park it and try a variation or three. The Tenori-On provides 16 'blocks', each being a complete set of 16 layers. You can switch instantly between any of these blocks with the R5 key, and with practice, you learn to make this part of your performance, grabbing a few notes from one block, then switching and playing a few notes from another. It's a dynamic means of song creation, and because you will be doing this sort of thing often, block and layer copy functions are conveniently provided — also via the R5 key.
It works like this: When you hold down R5, the current block is represented as a lit vertical column. Each time you push a button in the column its intensity changes, toggling between three options: switch block, copy layer or copy entire block. This means you can avoid the menu system for those operations you'll perform regularly. Neat.
One function I looked for but failed to find was 'solo layer' — then I discovered the workaround of copying the current layer into an empty block (assuming you have one). Voilà — instant solo. But there were a few limitations I didn't overcome so easily. For example, the settings that you make for one block affect all of them. Choosing one sound per layer makes sense, but when transposition, layer speed, note length and so on prevail across all notes and all blocks too, this seems too restrictive. I really wanted to set up a series of transposed blocks for fast selection during performance, or even to define blocks using different scales, but could find no way to accomplish this other than to perform the changes manually every time. And no matter how fast you become with practice, you can still only adjust one parameter at once. Nevertheless, blocks are incredibly useful in performance.
The Tenori-On can communicate with other gear, courtesy of the 'Mini DIN' cable provided. Each of its layers transmits on a fixed MIDI channel, so Layer 1 transmits on channel 1, Layer 2 on channel 2 and so on. Actually, the Tenori-On suffers from a few shortcomings as a hardware sequencer. For example, it sends all its notes at a fixed velocity and sends few useful MIDI data types other than volume and pan controllers, plus program changes. In Score mode, up to 16 notes may be transmitted at once per layer over a range of C4 to Eb5 (MIDI notes 60 to 75). This is subject to transposition and scale correction.
By setting synchronisation to 'slave', you can sync the Tenori-On to your MIDI rig, which I found great for the resulting lightshow. Like an analogue sequencer, you are free to start and stop it manually when sync'ed, although if you do this, it takes practice to line up playback reliably.
In general, I suspect the MIDI interface will be used to transmit any ideas you generate into your favourite sequencer for further work, or for arrangement using alternate sound modules.
The built-in tone generator is based on Yamaha's sample-based AWM2 technology. Its maximum 32 notes of polyphony will run out long before the sequencer's ability to generate simultaneous notes. There are 253 preset voices over which you can exercise only the most basic control. Actually, you can vary the note length, volume and pan and that's your lot.
Some of the pads and blippy noises are undeniably cute, but it seems that creating quality bass patches was a low priority, perhaps due to the size of the built-in speakers. Voice names such as Slowflux, Ion, Photon, Raindrop and Helix suggest a good deal more exotica than they deliver, and while there are several piano and string voices, and lots of organs, many sound more home keyboard than professional synthesizer. There are 14 drum kits included, but even these are rather lightweight. Ultimately, I found the sonic character was blandly pervasive — as I realised when I started to compare my own fledgling songs with some of the demo MP3s retrieved from the net.
To add a little gloss, the Tenori-On offers a basic effects implementation consisting of reverb and chorus/flanger. These act as a light spray over all layers equally, which possibly explains why there's no delay: things could get messy. With a mere two parameters to adjust, type and depth, the effects are never obtrusive. Even so, for evolving, ambient pieces, I found the hall reverb to be fairly effective with its depth cranked up to maximum. In the end, all you can really say is that the effects are an innocuous, uncomplicated addition to the Tenori-On's sonic output.
In addition to the 253 preset patches, a further three are available for user 'samplings' — an alternate term for the more familiar 'multisample'. In the case of the Tenori-On, each user sampling consists of 16 separate samples of up to 0.97 seconds. These samplings are created using the supplied User Voice Manager software, which is a simple utility with only rudimentary file handling, but sufficient to do the job. I quickly prepared my three user voices, assembling them from various percussion, vocal and flute samples chosen at random. The software supports 8- or 16-bit stereo samples in WAV or AIFF format and a selection of sampling rates, including 44.1kHz and 48kHz. As it ignores any part of the sample beyond 0.97 seconds, the whole selection process is totally painless. I copied my three voices onto an SD card — the Tenori-On's way of storing and loading data — then popped this into the slot and set about loading. It's fortunate that user samples are maintained after power-off, because it took over two and a half minutes to load each one!
I was pleasantly surprised to hear how well my samples translated and what a difference this made to making music on the Tenori-On. I'm probably biased, but I'm convinced my own three voices — created with barely a synaptic burp on my part — were sonically superior to any of the supplied synth patches or drum kits. They were certainly louder. I began pondering what wonders could be achieved if only it were possible to replace the other 253 voices with user samples too!
One last thing to be emphasise concerning user voices is that they, in common with the drum kits, are not subject to any transposition or scale correction. So if you want to play tunes with them, you need to load in samples corresponding to each note you will need.
Leaving cost aside for a moment, there's much to admire about the Tenori-On. Like a fully evolved Electroplankton, it challenges us to think differently about how music can be made and its clarity of vision helps it quickly become intuitive, bordering on the addictive. For musicians it could be anything from a neat sketchpad, an ideas generator or a quirky performance instrument. For those with no musical background, it could be a fun introduction to creating tunes and noises.
The light show is undeniably lovely and I could easily imagine spending many hours hypnotised by it when under the influence of something slightly stronger than coffee. Drawing musical patterns using the various modes produces results you wouldn't obtain any other way, making it doubly fortunate that its MIDI output can be routed into your sequencer.
If only you could replace all of its sounds with original samples, or if you had timbral control over the existing voices, or if it offered more sophistication as a hardware sequencer, the Tenori-On could cast aside any toylike comparisons. While its limitations could be ignored in a cheaper gadget, I'm afraid they do stand in the way of a good time at the asking price. This is a pity because we live in a world of increasingly conservative, neutered, more-of-the-same products, and Yamaha should be applauded for making Toshio Iwai's dream come true and daring to release the most unusual little musical instrument of recent years. Let's hope this isn't the end of the story...